CES 2026 is officially open in Las Vegas, and the show floor already makes one thing painfully clear: the AI boom has gone fully physical.
From data center racks the size of two compact cars to cute R2D2-style bots and a robotaxi built on a luxury SUV, this year’s CES is where silicon, software and hardware finally collide.
TechCrunch reporters led by Kirsten Korosec and Sarah Perez are on the ground, filing a steady stream of live updates. Here’s what’s standing out from the first official day.
AI is for everyone — if you can get the compute
AMD chair and CEO Dr. Lisa Su delivered the first official CES keynote, and she didn’t bother to pretend this show is about anything but AI.
“AI is the most important technology of the last 50 years, and I can say it’s absolutely our number one priority at AMD,” she said. “It’s already touching every major industry, whether you’re going to talk about health care or science or manufacturing or commerce, and we’re just scratching the surface, AI is going to be everywhere over the next few years. And most importantly, AI is for everyone.”
Su backed that up with adoption numbers. Since the launch of ChatGPT, she said we’ve gone from about a million people using AI tools to more than a billion active users — a ramp that took the internet decades to hit. AMD now projects AI adoption will grow to over 5 billion active users.
The catch? Compute.
Su’s keynote hammered home that everything from healthcare to national security will be constrained by access to GPUs and supporting infrastructure.
OpenAI president and co-founder Greg Brockman joined her on stage and more or less confirmed that.
“Every time we want to release a new feature, want to produce a new model, we want to bring this technology to the world, we have a big fight internally over compute because there are so many things we want to launch and produce for all of you that we simply cannot, because we are compute constrained,” he said.
He went further, arguing that GDP growth itself may soon be tied to how much compute a country controls.
“I think that we’re starting to see the first inklings of this,” Brockman said. “And I think over the next couple of years we’ll see it start to hit in a real way.”
Brockman also repeated his now-familiar wish list item:
“The world is going to require far more compute than we have right now. Like, I would love to have a GPU running in the background for every single person in the world, because I think it can deliver value for them, but that’s billions of GPUs. No one has a plan to build that kind of scale.”
AMD’s answer: heavier racks, smarter PCs
To make a dent in that compute gap, AMD came to CES with both infrastructure and PC silicon.
On the data center side, Su rolled out Helios, an open, modular rack design first revealed in 2025 and back on stage this year. It’s built on the OCP Open Rack Wide standard in collaboration with Meta.
“Helios is a monster of a rack,” Su said. “This is no regular rack, okay. This is a double wide design based on the OCP open rack wide standard developed in collaboration with Meta, and it weighs nearly 7,000 pounds.”
Or as she put it more bluntly:
“We wanted to show you what is really power in all this AI, it is actually more than two compact cars.”
Closer to the edge, AMD is betting that AI-powered PCs will be the next big upgrade cycle. Su announced the Ryzen AI 400 Series, the latest generation of AMD’s AI PC chips.
According to AMD’s own numbers, the new processors deliver 1.3x faster multitasking and are 1.7x faster for content creation than rival chips. The pitch: on-device AI assistants, creative tools and automation that don’t have to round-trip to the cloud for every task.
Nvidia’s cute robots and a battle for the full stack
Nvidia CEO Jensen Huang used his keynote to reiterate the company’s ambition to own the entire AI stack, from chip architectures like Rubin to software and training models such as Alphamayo.
But it was the robots that stole the show.
Two R2D2-like bots waddled slowly across the stage — making their second appearance on an Nvidia keynote — as Huang talked through the company’s robotics roadmap. He’s previously said he believes humanoid robotics could become a multi-trillion-dollar industry, and the demo was a small, charming hint at where Nvidia wants to go.
“Hurry up,” he joked, as they trundled across the stage.
Nvidia also looms large behind the scenes of the broader robotics wave sweeping CES, from factory automation to autonomous vehicles.
World-building AI: Fei-Fei Li’s Marble
Dr. Fei-Fei Li, co-founder and CEO of World Labs, joined Su on stage to show how generative AI isn’t just about text and images.
World Labs’ first product, Marble, generates 3D worlds that still obey the laws of physics.
Marble can spin up imaginary environments from scratch or reconstruct real places — like parts of Las Vegas — from photos into explorable 3D scenes.
“What you don’t see here behind the scene is how much computation and why inference speed really matters,” Li said. “The faster we can run these models, the more responsive the world becomes. Instant camera moves, instant edits and a scene that stays coherent as you actually navigate and explore and that’s what’s really important.”
It’s a neat example of the broader CES 2026 theme: AI leaving the browser tab and reshaping the physical and spatial worlds we move through.
Lucid, Nuro and Uber show their robotaxi
On the transportation side, Uber, Lucid and Nuro finally pulled the wraps off the production-intent version of their robotaxi.
Built around the Lucid Gravity SUV, the vehicle feels overwhelmingly spacious inside — a sensible choice for a premium robotaxi play. The rider UI looks a lot like Waymo’s in-car interface, but Uber says it’s building the software itself.
If everything holds, the trio plans to start offering rides in the Bay Area later this year, positioning the Gravity-based service as a roomier alternative to Waymo’s Jaguar I‑Pace fleet.
The bots invade the office and the living room
Beyond the big-name keynotes, the show floor is already producing its usual stream of very CES hardware.
Vibe’s meeting-room bot
Vibe, an HSG-backed startup (formerly Sequoia China) known for its collaborative digital whiteboards, launched a physical meeting assistant bot.
The device:
- Packs a 4K camera that can track attendees and double as a webcam
- Uses beamforming microphones to take meeting notes
- Includes a voice assistant and a touchscreen to display information
- Integrates with calendars and project management tools so you can set reminders, delegate tasks and review past meetings
The price: $1,799 at launch.
This is yet another example of AI agents jumping off the laptop screen and into dedicated hardware.
A robot vacuum that perfumes your home
Over in smart home land, Anker’s Eufy brand is trying to differentiate in a crowded robot vacuum market with the Clean Robot Vacuum Omni S2, priced at $1,600.
The S2 doesn’t just vacuum and mop — including on longer shag carpets — and clean its own mop head with sanitized water (the tank electrolyzes the water, turning it into hypochlorous acid).
It also dispenses home fragrance as it cleans, something Eufy says no one else is doing yet.
Lift the lid and you’ll find a chamber for scent pods: solid gels that slowly release fragrance as the robot runs. The unit ships with three different scents, and refills will be available.
And we’re just getting started
All of this comes on top of a pre-show news dump that included:
- Nvidia’s broader AI announcements
- AMD’s expanding chip lineup
- A strategic partnership between Google DeepMind and Boston Dynamics
- A packed AMD keynote roster featuring OpenAI’s Greg Brockman, Fei-Fei Li of World Labs, Absci CEO Sean McClain, Generative Bionics co-founder Daniele Pucci, John Couluris of Blue Origin and more
The Las Vegas Convention Center may have only just opened its doors for CES 2026, but the through-line is already obvious: compute capacity, not imagination, is the limiting factor for the next wave of AI.
Expect the rest of the week to keep testing that boundary — one robot, rack and robotaxi at a time.
